Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A Monte Carlo simulation based on first-passage distributions for spatio-temporal detection on potentiometric sensor arrays.

Analytical and bioanalytical chemistry·2026
Same author

Identifying Actionable Glucose Patterns in Type 1 and Type 2 Diabetes Using Continuous Glucose Monitoring and Algorithmic Pattern Recognition.

Journal of diabetes science and technology·2026
Same author

Glucose Predictions Improve Glycemic Control: A Digital Twin Evaluation.

Diabetes technology & therapeutics·2026
Same author

Efficacy of an AI-Enabled Low Glucose Prediction: A Pooled Performance Analysis With Capillary Blood Glucose as Ground Truth.

Journal of diabetes science and technology·2026
Same author

Artificial intelligence for improving decision-making in bacterial infection management: a narrative review.

The Journal of antimicrobial chemotherapy·2025
Same author

The impact of artificial intelligence-driven decision support on uncertain antimicrobial prescribing: a randomised, multimethod study.

The Lancet. Digital health·2025

Related Experiment Video

Updated: Oct 5, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.4K

A Deep Learning Framework for Automatic Meal Detection and Estimation in Artificial Pancreas Systems.

John Daniels1, Pau Herrero1, Pantelis Georgiou1

  • 1Centre for Bio-Inspired Technology, Department of Electrical and Electronic Engineering, Imperial College London, London SW7 2AZ, UK.

Sensors (Basel, Switzerland)
|January 22, 2022
PubMed
Summary

A new deep learning algorithm automates meal detection and carbohydrate estimation for artificial pancreas (AP) systems. This improves glucose control in type 1 diabetes (T1D) management without increasing hypoglycemia risk.

Keywords:
artificial pancreascarbohydrate estimationdeep learningmachine learningmeal detectionmultitask learningneural networkquantile regressiontype 1 diabetes

More Related Videos

Concept Development and Use of an Automated Food Intake and Eating Behavior Assessment Method
06:21

Concept Development and Use of an Automated Food Intake and Eating Behavior Assessment Method

Published on: February 19, 2021

5.9K
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

4.3K

Related Experiment Videos

Last Updated: Oct 5, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.4K
Concept Development and Use of an Automated Food Intake and Eating Behavior Assessment Method
06:21

Concept Development and Use of an Automated Food Intake and Eating Behavior Assessment Method

Published on: February 19, 2021

5.9K
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

4.3K

Area of Science:

  • Biomedical Engineering
  • Artificial Intelligence in Medicine
  • Endocrinology

Background:

  • Current artificial pancreas (AP) systems require manual meal announcements, posing a burden for type 1 diabetes (T1D) management.
  • Effective postprandial glucose control is challenging due to reliance on frequent user engagement.

Purpose of the Study:

  • To develop and evaluate a deep learning algorithm for automated meal detection and carbohydrate estimation.
  • To enhance AP systems for fully automated closed-loop glucose control.

Main Methods:

  • A deep learning framework utilizing multitask quantile regression was developed.
  • The algorithm was evaluated in silico using the UVa/Padova simulator with 10 adult subjects and a Bio-inspired Artificial Pancreas (BiAP) control algorithm.
  • Three AP configurations were compared: BiAP without meal announcement (BiAP-NMA), BiAP with meal announcement (BiAP-MA), and BiAP with meal detection (BiAP-MD).

Main Results:

  • BiAP-MD demonstrated improved glucose control compared to BiAP-NMA, with a lower mean blood glucose level (-4.4 mg/dL, p<0.01) and increased time in range (+3.9%, p<0.001).
  • The meal detection algorithm achieved 93% precision and 76% recall, with a detection delay of 38 ± 15 minutes.
  • BiAP-MD showed better hypoglycemia management than BiAP-MA, with fewer control errors (10% vs. 20%).

Conclusions:

  • Multitask quantile regression effectively improves AP systems' postprandial glucose control capabilities.
  • Automated meal detection and carbohydrate estimation enhance AP system performance without increasing hypoglycemia.
  • This approach moves towards fully automated closed-loop glucose control for T1D management.